Professional Documents
Culture Documents
719
PS70CH30_Scherer ARI 9 November 2018 9:13
Contents
INTRODUCTION AND SCOPE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 720
COMMON GROUND AMONG EMOTION THEORIES . . . . . . . . . . . . . . . . . . . . . . . 721
AN APPROACH TO ORGANIZING EMPIRICAL WORK REGARDING
THE EMOTION PROCESS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 722
METHODS FOR THE MANIPULATION AND MEASUREMENT
OF COMPONENTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 723
Manipulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 723
Measurement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 724
PATH A: FROM APPRAISAL TO ACTION TENDENCIES
AND BEHAVIOR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 724
Access provided by PERI - ET - Addis Ababa University on 08/09/21. For personal use only.
of action readiness or action tendencies, and physically, in the form of physiological responses;
Annu. Rev. Psychol. 2019.70:719-745. Downloaded from www.annualreviews.org
(c) the integration of information acquired from these two steps into a central representation that
allows monitoring and regulation of the potential responses; and (d ) often, although not always,
the categorization and communication of the emotion episode to other people.
Imagine the following situation: During a party, you happen to overhear two of your friends
saying unpleasant things about you. Would this elicit an emotional reaction in you, and if so,
which one? In response to this question, in a sample of more than 3,000 working adults in dif-
ferent countries, 42.8% indicated that they would feel anger, 23.8% contempt, 0.7% fear, 14.2%
good humor, 2.3% guilt, 38.3% sadness, 4.0% shame, and 5.0% worry (K.R. Scherer, unpub-
lished manuscript). Responses were obtained with a forced choice list of emotions. In many cases,
respondents indicated more than one emotion, suggesting the experience of mixed or blended
emotions. Most likely, with open response options, an even larger range of different emotional
reactions would be obtained, including no emotional reaction at all. Ideally, psychological sci-
ence should provide a theoretical framework that predicts for whom an emotion episode will be
triggered in this situation (elicitation), what the nature of the reaction will be (differentiation),
what type of nonverbal feeling will be experienced (representation), and how it will be verbally
described (categorization, labeling).
Figure 1 illustrates the complete emotion process, in line with the definition of functions
and components proposed above. Emotions are not random events—they occur when something
happens to us. In the elicitation stage, triggers of emotions can be objects (visual, verbal, olfactory),
acts of nature, the behavior of others, and our own actions. Possible elicitors can also be internal
processes, such as imagination or memories of events, hormonal changes or drug effects, and
voluntary decisions to experience certain emotions. The elicitors impacting our organism are
filtered by our sensory organs and subject to intensive information processing, most often through
multilevel appraisals (e.g., in the example discussed above, “Do they really mean this? Did I do
something wrong? What will this mean for our relationship? What can I do?”).
The appraisal results produce differentiation—various action tendencies and corresponding
physiological responses, motor expressions, and feelings (e.g., a tendency to aggress or avoid,
an increase or decrease of muscle tonus, brow lowering or raising, and feeling hot or cold).
The continuous changes in these mental and bodily components are integrated and centrally
represented as feelings (in nonverbal form) (see Moors 2017, Scherer 2009a); these responses can
be categorized and labeled with emotion terms (e.g., anger, fear, sadness) or descriptions referring
to dimensions (e.g., feeling good or bad, feeling aroused or calm) (see Scherer 2001, 2009a).
Most of the relevant research to date has measured feelings with self-report via rating scales with
emotion or dimension labels. Given the difficulty of the task, there have been virtually no attempts
to measure the nonverbal content of the central representation.
Action tendencies
A
• Event
• Behavior B D E • Categorization
Physiological Component integration—
• Situation Multilevel appraisals
responses experienced feeling • Labeling
• Object
• Memory
C
Motor expression
Access provided by PERI - ET - Addis Ababa University on 08/09/21. For personal use only.
Annu. Rev. Psychol. 2019.70:719-745. Downloaded from www.annualreviews.org
Figure 1
The dynamic architecture of a multicomponent emotion process model reflecting the widely held assumption that emotion episodes are
processes that are elicited by a cognitive evaluation or appraisal of events, producing synchronized changes in several components. The
hypothetical model illustrates the assumed mechanism. While the initial causal effects move from left to right, the assumption is that
there is a high degree of recursiveness, i.e., the result of the initial impact of one component on another is expected to feed back to the
eliciting component, setting off another round of processing likely to affect the ultimate outcome. This is why most of the arrows in the
model are bidirectional, as feedback effects have already been shown or are highly probable. This article focuses on the empirical
evidence for the hypothesized effects of initial evaluation results, categorized by the major appraisal criteria described in the literature,
on major response components—motivational action tendencies (A), physiological reactions (B), and motor expressions (C). It is
assumed that these response components also interact among each other. The conjoint effects of these interactions are continuously
represented in central regions of the brain, likely to become accessible to consciousness in the form of nonverbal feelings (D). These
feelings can consequently be categorized and labeled with emotion words or verbal emotional expressions (E). Due to space restrictions,
effects D and E are not discussed in the article. Figure modified with permission from Scherer (2009a).
Most emotion theorists do not fundamentally disagree about the emotion process as concep-
tualized in Figure 1, but they differ in the components on which they focus. In addition, theories
that focus on similar components differ with regard to the details of the mechanisms involved.
For instance, theories postulating that some kind of appraisal initiates the emotion process tend
to assign different weights to the various appraisal criteria and thus have put forward different
hypotheses regarding the influence of these criteria on the other components of the emotion
episode. Proposals range from template-based stimulus mapping, as in simple stimulus–response
models, to attribution models (focusing on the attribution of stimuli to different types of causes),
social judgment models (focusing on social relationships), decision theories (focusing on values
and expectancies of action alternatives), and appraisal theories (proposing a set of appraisal criteria
or dimensions such as goal relevance, valence, control, agency, and fairness) (for reviews of various
theories, see Moors 2009, 2017; Scherer 2009b; Scherer & Peper 2001). The fact that different
theories have different focuses does not necessarily make them incompatible. The remainder of
this review does not focus on potential differences. Rather, it presents a summary of empirical
research that speaks to a variety of theoretical predictions in the literature.
a componential view, many others continue to study emotions by focusing on emotion categories,
thereby neglecting to address the entire emotion process and the relationships among components.
Nevertheless, scanning the literature reveals a substantial body of research on these relationships.
Most of this research, however, has studied the relationship between only two components at a
time, with the exception of a few studies that have studied several components simultaneously
(Bossuyt et al. 2014, Gentsch et al. 2014).
For this review, three paths have been selected: from appraisal to action tendencies, from
appraisal to physiological responses, and from appraisal to motor expression (facial, vocal, and
gestural). The choice to include appraisal in each of the relationships is based on the postulate,
widely shared among emotion theories, that it is the nature of the appraisal process that determines
the nature of the remaining stages in the emotion process. For each relationship, research is
considered for each of five major appraisal criteria (see Ellsworth & Scherer 2003): novelty or
Access provided by PERI - ET - Addis Ababa University on 08/09/21. For personal use only.
expectedness, valence (intrinsic and extrinsic), agency, control or power, and fairness. Many studies
have focused on the isolated influence of one of these criteria; some have addressed the influence
Annu. Rev. Psychol. 2019.70:719-745. Downloaded from www.annualreviews.org
of patterns of two or more. In this review, we focus on initial appraisal after the occurrence of an
event and do not address the copious literature on reappraisal as part of emotion regulation.
In experimental research about relationships among components, one component must be
manipulated, and the other component must be measured. It is useful to discuss the methods for
manipulation and measurement up front and highlight some of their advantages and disadvantages.
Manipulation
Manipulation of overt bodily components can be done directly or indirectly (see Moors & Scherer
2013). For instance, physiological responses can be produced directly by injection of chemical
substances or indirectly via behavior (e.g., running up the stairs leads to a rise in heart rate).
Likewise, facial muscles can be activated directly via electrical current or indirectly via instructions
to tense the muscles in particular ways.
Mental components can, in principle, only be manipulated indirectly, but also in this case, some
methods are more indirect and others are more direct. Appraisals of events can be manipulated
indirectly by presenting real events (e.g., insulting a participant) or virtual events (e.g., pictures,
films, words, stories) that the researcher expects to be evaluated in a particular way. They can also
be manipulated by asking participants to imagine or recall these events (e.g., recall an episode in
which you were insulted). These methods are indirect in that the researcher makes no mention
of the appraisal criteria under study. More direct methods ask participants to imagine or recall
events described in evaluative terms (e.g., imagine or recall a negative event in which you had low
control) (Fast & Chen 2009, Galinsky et al. 2003, Kuppens et al. 2003, Lammers et al. 2008) or
present the evaluative terms themselves to prime particular appraisals (Smith & Bargh 2008). Ma-
nipulating appraisals with indirect methods has the advantage that the researcher can control the
nonevaluative features of the event. The more direct methods, however, have the advantage that
the researcher can control the evaluative features of the event. All methods, except imagination
and recall, have the further advantage that they allow manipulating the appraisal in a fairly unob-
trusive way, thereby diminishing the risk for demand, as well as the risk for activating participants’
convictions about relationships between appraisals and other components (Parkinson 1997).
Action tendencies, which are also considered to be mental entities, are manipulated indirectly
via instructions to act in certain ways (e.g., to approach or avoid) (Kozlik et al. 2015). The feeling
negative response)
from another variable that is assumed to influence the to-be-measured variable (De Houwer &
Annu. Rev. Psychol. 2019.70:719-745. Downloaded from www.annualreviews.org
Moors 2010). For instance, behavioral choice can be taken as a direct measure of behavior and
as an indirect measure of action tendencies (Martinez et al. 2011), based on the assumption that
action tendencies influence behavior. The combination of both dichotomies results in four types
of methods: direct and indirect objective methods and direct and indirect subjective ones.
Overt bodily components can, in principle, be measured with all types of methods, but are best
measured by direct objective ones. For instance, physiological responses are sometimes reported by
the person having them (i.e., direct subjective), but more often they are measured with technical
devices for registering heart rate, skin conductance, and blood pressure (i.e., direct objective).
Facial expressions are measured with direct objective methods such as the facial action coding
system (FACS; Ekman & Friesen 1978), in which observers code visible facial action units (AUs)
(i.e., actions of parts of the face), or with electromyography (EMG), a technique that registers
visible and invisible facial muscle movement in an area of the face. Vocal expressions can be
measured by acoustic analysis of voice and speech ( Juslin & Scherer 2005).
Mental components can be measured with all types of methods except direct objective ones. Ap-
praisals of events and central representations or feelings have been measured with self-report (i.e.,
direct subjective), but also with stimulus–response compatibility tasks relying on reaction times
(i.e., indirect objective). Action tendencies have likewise been measured with self-report (Frijda
et al. 1989) and stimulus–response compatibility tasks (i.e., indirect objective) (Kozlik et al. 2015).
With these methods in mind, the following three sections review empirical research pertaining
to paths A, B, and C of the model shown in Figure 1. Paths D and E are not discussed due to space
limitations and lack of a coherent body of empirical literature guided by theoretical predictions.
earlier observations. Novel or unexpected as well as goal-relevant stimuli lead to a shift in attention,
which can be considered a mental action tendency (e.g., Brosch et al. 2013).
are instructed to approach positive stimuli and avoid negative ones on compatible trials, and to
Annu. Rev. Psychol. 2019.70:719-745. Downloaded from www.annualreviews.org
approach negative stimuli and avoid positive ones on incompatible trials. Responses are typically
faster and less error prone on compatible than on incompatible trials. This effect has been
obtained across a variety of response modalities, such as pulling or pushing a lever, approaching
or avoiding a manikin on a computer screen, nodding or shaking the head, and stepping forward
or backward (for a review, see Kozlik et al. 2015). The effect has been interpreted as evidence that
valenced stimuli activate action tendencies, which facilitate similar instructed action tendencies
(Krieglmeyer et al. 2013).
Alternative accounts for the link between intrinsic valence and approach or avoidance have
been proposed. Eder & Rothermund (2008) argued that compatibility effects are not evidence that
valenced stimuli spontaneously activate action tendencies. Instead, these effects could be produced
by semantic overlap between the valence of the stimuli and the valence of the instructed responses.
Other studies show that negative stimuli do not always lead to an avoidance tendency but sometimes
to an aggressive tendency, and that the latter can be seen as one specific type of approach tendency
(Berkowitz & Harmon-Jones 2004, Krieglmeyer & Deutsch 2013). However, if approach and
avoidance tendencies are conceptualized on a higher level of abstraction, as tendencies to increase
and decrease physical or social contact, the tendency to aggress can be viewed as a strategy for
ultimately decreasing contact (Frijda 1986).
Negative stimuli lead not only to an avoidance tendency, but also to a tendency to repair. This
is supported by research on cognitive control and conflict monitoring (Botvinick 2007, Dignath
et al. 2015). Studies in this area show that conflict stimuli (i.e., stimuli that call for two competing
actions, such as Stroop stimuli), which are incongruent with the goal to perform well on the task,
instigate the tendency to adjust one’s behavior (i.e., put in more effort to overcome the conflict)
and activate the tendency to avoid subsequent conflict.
Effect 1: agency or blame + valence → tendency to aggress. Early research showed evidence
for higher aggressive tendencies when a person is blamed more for the occurrence of a negative
event (e.g., Kulik & Brown 1979; for a review, see Berkowitz & Harmon-Jones 2004). Recent
studies have investigated reappraisal, rather than first-time appraisal, of agency (for a meta-analysis,
see Barlett 2013). Participants are first led to believe that another person caused them harm (other-
agency), after which they receive mitigating information indicating that the person did not cause
the harm or was not responsible for it (circumstances-agency). These studies show a reduction
in aggression after the presentation of some but not all types of mitigating information (see also
Barlett et al. 2016).
Effect 2: agency + valence → tendency to repair. One line of research examines the influence
of self- or other-agency on the tendency to repair when the negative event happened to another
person. In an early lab study by Carlsmith & Gross (1969), participants who caused harm (shocks) to
another person (self-agency) were more likely to comply with an unrelated request than those who
Access provided by PERI - ET - Addis Ababa University on 08/09/21. For personal use only.
observed someone else cause the harm (other-agency). A field study by Konecni (1972) obtained
Annu. Rev. Psychol. 2019.70:719-745. Downloaded from www.annualreviews.org
the opposite pattern of results. Participants who caused harm themselves (self-agency) repaired
less but avoided more than participants who watched someone else cause it (other-agency). Later
studies focused on the role of guilt feelings as a mediator between self-agency appraisal and repair
behavior, with some studies showing evidence in favor of (Nelissen et al. 2007) and others evidence
against this interpretation (Parkinson & Illingworth 2009).
Another line of research examines the influence of self- or nonself-agency on the tendency to
repair when the negative event happened to the participant. The evidence is mixed in this case as
well. Zeelenberg & Beattie (1997) showed, using an ultimatum game paradigm, that self-agency
combined with a negative stimulus leads to regret and that more regret leads to a stronger tendency
to repair. A different result was obtained by Bossuyt et al. (2014), who used a computer game to
manipulate agency regarding a negative game event. They showed stronger feelings of regret but
not a stronger tendency to repair in the self-agency compared to the circumstances-agency trials.
A subsequent recall study showed that control, rather than agency, is crucial for the tendency
to repair and suggested that previous studies found an effect of agency because they confounded
agency with control.
In the following sections, we discuss empirical findings in favor of and against four effects: the
influence of general and prospective control on (a) the tendency to aggress or engage in other
antisocial behavior (Effect 3), (b) the tendency to approach or avoid (Effect 4), (c) the tendency to
be active or passive (Effect 5), and (d) the influence of control in the sense of a specific action (e.g.,
one can escape) on the tendency to engage in that action (e.g., to escape) (Effect 6).
Effect 3: high (low) control (+ valence) → more (less) aggressive or other antisocial ten-
dencies. This section separately considers data for general and prospective control. The strongest
evidence for the influence of high general control on the tendency to aggress (and other forms
of antisocial behavior) comes from research on social power, in which high-power participants
were more likely to act aggressively, as manifested in familial aggression, crime against minorities,
sexual harassment, bullying, and corruption (Guinote 2017).
Access provided by PERI - ET - Addis Ababa University on 08/09/21. For personal use only.
These effects are typically explained as mediated by the activation of neural circuits known
as the behavioral approach system (BAS) and the behavioral inhibition system (BIS) (Gray &
Annu. Rev. Psychol. 2019.70:719-745. Downloaded from www.annualreviews.org
McNaughton 1996). High power both increases activation of the BAS, leading to a stronger
tendency to approach one’s goals, and decreases activation of the BIS, leading to less inhibition of
aggressive strategies to reach these goals. Both elements contribute to increased aggression. Low
power, in contrast, increases activation of the BIS, resulting in decreased aggression.
A different set of studies suggests that people high in power engage more in prosocial than in
antisocial behavior (Chen et al. 2001, Fleischmann et al. 2018). In an attempt to reconcile these
contrasting findings, Guinote (2017) argued that high power increases the pursuit of salient goals.
Various moderators (e.g., task demands, organization and national culture, and personality traits
and needs) determine whether the salient goals are selfish or prosocial. Selfish goals lead to more
corrupt and antisocial behavior, whereas prosocial goals lead to more social behavior.
The hypothesis that high and low prospective control over a negative event should, respec-
tively, lead to the tendencies to aggress and avoid has intuitive appeal (Ellsworth & Scherer
2003, p. 580), yet experimental evidence for it is scarce (see Berkowitz & Harmon-Jones 2004).
However, to the extent that, in studies in which general control (social power, self-efficacy) was
manipulated, this manipulation also influenced participants’ prospective control, these studies
can be considered as indirect support for the link between high (low) prospective control and the
tendency to aggress (avoid).
Effect 4: high (low) control → tendency to approach (avoid). The observation that high
control sometimes leads to antisocial and sometimes to prosocial tendencies, and that both of
these tendencies can be captured under the denominator of approach tendencies, has led some
researchers to propose that the relationship between control and action tendencies is best described
as a relationship between high (low) control and approach (avoidance) (Guinote 2017), which is
consistent with the BAS and BIS explanation.
Several findings speak more directly to this relationship. For instance, Smith & Bargh (2008)
manipulated power by using a priming procedure. Their results revealed that high-power partic-
ipants had a stronger tendency to approach than low-power participants and non-power-primed
participants. The groups showed no difference in the tendency to avoid.
Effect 5: high (low) control + negative valence → tendency to be active (passive). Several
effects that have been described as effects on the tendency to approach (avoid) could be reinter-
preted as effects on the tendency to be active (passive) because, in many studies, the approach
behaviors were more active than the avoidance behaviors (but see Smith & Bargh 2008 for an ex-
ception). A number of research findings yield more direct support for the link between control and
activity (or the intensity of action tendencies). This section separately considers data for general
control, prospective control, and the combination of both types of control.
Early studies in the learned-helplessness literature established that subjects (animals, humans)
confronted with a series of uncontrollable aversive stimuli in a training phase (unavoidable shocks
or loud noise, unsolvable puzzles) showed decreased effort leading to impaired performance in
a test phase compared to subjects who were trained with controllable stimuli (e.g., Hiroto 1974,
Seligman 1975). This can be interpreted as evidence that high (low) general control leads to the
tendency to be more (less) active. More recent findings from the achievement motivation literature
are in line with these results. Students or workers often exhibit more persistence, resulting in better
performance, when they have high (compared to low) self-efficacy (for meta-analyses, see Multon
et al. 1991, Talsma et al. 2018) and when they have an internal (compared to an external) locus of
control (for a review, see Galvin et al. 2018).
Access provided by PERI - ET - Addis Ababa University on 08/09/21. For personal use only.
Further support for the link between high (low) general control and the tendency to be active
(passive) comes from research on social power. Galinsky et al. (2003) found that power holders
Annu. Rev. Psychol. 2019.70:719-745. Downloaded from www.annualreviews.org
have an increased action orientation, irrespective of whether this action was prosocial or antisocial.
The studies cited on learned helplessness, achievement motivation, and social power speak not
only to the influence of general control on the tendency to be active or passive. Insofar as the
manipulation of general control (or social power) in the training (or priming) phase transferred
to the stimuli in the test phase, these studies also provide evidence—albeit indirectly—for the
influence of high (low) prospective control on the tendency to be active (passive).
Research on achievement motivation provides more direct evidence for the latter link. In this
research, prospective control, or the expectancy to be able to perform a particular task, is typically
operationalized in terms of task difficulty—with easy (difficult) tasks considered as easy (difficult)
to control. Expectancy-value theories argue that the expectancy to succeed in a task interacts with
(i.e., is moderated by) the value attached to succeeding in the task (Nagengast et al. 2011): High
prospective control only leads to the tendency to be more active if the outcome of the action is
also desired.
Other studies, by contrast, showed a different pattern of results: Low prospective control did
not lead to passivity, but rather to reactance (resistance, opposition). This effect is also opposite
to Effect 3. In an early study, Donnerstein & Wilson (1976) observed that participants who had
to solve a math task under high-intensity noise displayed more aggression when they had low
than when they had high control over the noise. Wortman & Brehm (1975) reconciled these
contradictory results by arguing that, when participants initially experience a low amount of
failure, they still expect some control over the stimulus, which generates a reactive tendency. As
failures accumulate, they expect less control, which results in a passive tendency.
In some studies on achievement motivation, general and prospective control were manipulated
or varied orthogonally. A meta-analysis by Stajkovic & Luthans (1998) reported that participants
with high (low) self-efficacy had high (low) work performance for easy, medium, and difficult
tasks. Yet the effect was most pronounced for easy tasks, suggesting an additive effect of general
and prospective control on the tendency to be active. A somewhat different pattern of results was
obtained by Smith & Kirby (2009), who asked participants varying in subjective and objective math
ability to solve easy versus difficult math problems. They also found that higher-ability participants
had higher persistence and higher performance, but only for the difficult tasks. Importantly, the
effect was mediated by appraisals of higher prospective control (which they termed problem-
focused coping potential).
Effect 6: control in the sense of a specific action → tendency to engage in that action.
McCloskey et al. (2005) showed that having more options to control an aversive outcome does
not always lead to an aggressive tendency; sometimes it leads to an escape tendency (i.e., the
reverse of Effect 3). This is in line with older studies (e.g., Cherek et al. 1990) that show that
aggressive responses were maintained when they were rewarded by escape from point loss but were
Autonomic nervous
extinguished when they were no longer rewarded. This is in line with expectancy-value theory, system (ANS):
according to which the action tendency activated is in large part determined by the expected utility responsible for control
of the action (Moors 2017, Moors et al. 2017). of the bodily functions
that are not
consciously directed,
Appraisal Dimension: Fairness such as heartbeat
Most research relevant to the influence of fairness on action tendencies or behavior is conducted
with the ultimatum game (for reviews and meta-analyses, see Güth & Kocher 2014, van Damme
et al. 2014), in which a first player proposes a division of a sum of money between themselves and a
Access provided by PERI - ET - Addis Ababa University on 08/09/21. For personal use only.
second player. If the second player accepts this division, then the money is divided as indicated; if
they reject it, then neither player receives anything. A recurring observation is that second players
Annu. Rev. Psychol. 2019.70:719-745. Downloaded from www.annualreviews.org
reject unfair offers, causing harm to both players. Thus, this research documents the influence
of unfairness on the tendency to engage in costly aggression. Rejecting any division that delivers
more than zero has been labeled as irrational, at least if the (selfish) goal to maximize profit is the
only goal that is considered. If the (altruistic) goal for fairness is taken into account, however, then
rejecting unfair offers no longer qualifies as irrational.
The standard explanation for the observed effects is that receiving an unfair offer leads to
a negative emotion, which, in turn, leads to rejection (e.g., Matarazzo et al. 2016). Alternative
explanations leave out (full-blown) emotions as mediating entities (Bierbrauer et al. 2017, Civai
2013, Moors & Fischer 2018, Schank et al. 2017).
Bediou & Scherer (2014) manipulated both the goal incongruence and the fairness of offers
in an ultimatum game. They found that participants’ rejection rates of these offers were more
influenced by goal incongruence than by fairness.
emotion. She proposed that, rather than focusing solely on the final outcome (the emotion cate-
gory), researchers should place greater emphasis on emotion-antecedent information processing.
Early studies that measured the physiological effects of specific appraisal outcomes support the
notion that physiological activity associated with emotional reactions is organized around the
personal meaning attributed to the situation through an appraisal process (Pecchinenda 2001,
pp. 306–7).
The next sections discuss evidence on the effect of appraisal criteria—often mediated by action
tendencies—on physiological responses in the peripheral nervous systems.
massive changes in the ANS (e.g., an increase in muscular tone and skin conductance, a decrease
Annu. Rev. Psychol. 2019.70:719-745. Downloaded from www.annualreviews.org
in heart rate, pupil dilation, and a pause in respiration that is often followed by an increase in
respiratory depth and a decrease in respiratory rate) (Reisenzein et al. 2018). Likewise, Aue et al.
(2007) showed that highly goal-relevant (threatening) pictures increased heart rate more than did
neutral pictures. The orienting response serves to recruit extra sensory and cognitive resources
to identify and make sense of the source of the novelty or unexpectedness and the goal relevance
(Nieuwenhuis et al. 2011).
Delplanque et al. (2009) used intrinsically positive and negative odors and manipulated the
novelty of these odors to examine the effect of these appraisal criteria on several psychophysio-
logical parameters. They found a relative heart rate deceleration in response to novel odors and
a relative acceleration in response to negative ones, confirming the classical distinction between
orienting and defense reflexes.
the influence of threat versus challenge appraisals (based on perceived control and support) on
Annu. Rev. Psychol. 2019.70:719-745. Downloaded from www.annualreviews.org
cardiovascular reactivity (for recent examples, see Gaab et al. 2005, Gramer & Supp 2014, Harvey
et al. 2010; for reviews, see also Denson et al. 2009, Kemeny 2009).
Some studies combined appraisals of control and goal congruence (e.g., Chalabaev et al. 2009,
Pecchinenda & Smith 1996, Weinstein & Quigley 2006). Johnstone et al. (2005) studied par-
ticipants performing a computer task in which they could lose or gain points (goal congruence
manipulation) under two levels of difficulty (control manipulation). Skin conductance indicated
higher sympathetic arousal for losses than for gains, particularly when difficulty was high.
Facial Expression
Scherer et al. (2018) reviewed actor portrayal studies in which facial expressions of discrete emo-
tions were coded using the FACS (Ekman & Friesen 1978). Earlier predictions about emotion-
specific facial expressions (see Ekman 1997) were only partially confirmed. Many AUs were used
by actors as part of their portrayals of many different emotions. Mehu & Scherer (2015) also
showed that only a minority of facial AUs were associated with specific emotions. This is com-
patible with the hypothesis that appraisals combine in many different configurations, resulting in
a large variety of different emotions and emotion blends. Although the main rationale for actor
portrayal studies has been to find evidence for the existence of emotion-specific facial expressions
and the ability of judges to recognize these, it is possible to analyze the results of these studies for
evidence of this alternative hypothesis. Thus, the plausibility of the assumed functions of certain
facial expressions expected to be generated by specific appraisals can be investigated by considering
the links between the dominant appraisal configurations for major emotions (e.g., unexpectedness,
goal obstruction, and other-agency for anger) and the type of facial AUs consistently found for
these emotions. This approach was taken by Scherer & Ellgring (2007, pp. 125–6), who inter-
preted their empirical data on the facial expressions in actor emotion portrayals by discussing the
probability of certain appraisals being involved in generating these emotions.
More direct evidence for the hypothesis that appraisals drive specific facial actions can
be obtained when appraisals are manipulated in the laboratory and their effects on the face
are measured. Appraisals manipulated in the laboratory may not produce facial actions that can be
Access provided by PERI - ET - Addis Ababa University on 08/09/21. For personal use only.
detected through visual inspection, which is why EMG measurement is often used to measure the
innervation of specific muscles. A growing body of evidence documents specific muscle activation
Annu. Rev. Psychol. 2019.70:719-745. Downloaded from www.annualreviews.org
valence, and potency. A mediation model revealed that the association between facial behavior and
recognition of the signaler’s emotional intentions is mediated by perceived affective dimensions.
The authors concluded that the classification of emotional expressions into discrete categories
Somatic nervous
may therefore rely on the perception of affective dimensions and, presumably, their underlying system (SNS): also
appraisals. called the voluntary
Scherer et al. (2018) directly tested the claim that observers first detect specific appraisals from nervous system; the
different facial muscle actions and then use implicit inference rules to categorize and name specific part of the peripheral
nervous system
emotions. They reported three experiments in which various configurations of facial AUs expressed
associated with the
by synthesized avatars were judged for appraisals or emotions, and their results confirmed that par- conscious, voluntary
ticipants can, indeed, infer appraisals and emotions in line with predictions from appraisal theories. control of body
movements via skeletal
muscles
Vocal Expression
Access provided by PERI - ET - Addis Ababa University on 08/09/21. For personal use only.
Fundamental
Empirical work on emotional expression has been dominated by study of the face until very frequency (f0): the
Annu. Rev. Psychol. 2019.70:719-745. Downloaded from www.annualreviews.org
lowest frequency of a
recently. This is quite surprising, as there is an extraordinary amount of evolutionary continuity and
periodic waveform
interspecies similarity (at least for mammals) in the vocal channel of emotion expression (Briefer with several
2012). As is the case for facial expression, much of the work on the human voice has been fairly harmonics, e.g., the
atheoretical, basically attempting to identify the vocal profiles that differentiate specific emotions. voice, heard as pitch
Scherer (1986) proposed a theoretical framework to predict emotion-specific vocal expressions
as the result of a sequence of appraisals likely to result in action tendencies and physiological [ANS
and somatic nervous system (SNS)] changes that could affect vocalization in a specific fashion. The
following examples illustrate this approach and are accompanied by available empirical evidence
for several appraisal criteria.
Appraisal criterion: novelty or expectedness. The orienting response (see above) is likely to
produce corresponding changes to the settings of the extrinsic and intrinsic laryngeal muscles,
leading to raised fundamental frequency (f0; heard as pitch) due to the raised vocal fold tension
and respiratory changes affecting voice intensity and speech fluency.
stimulus (see the section titled Path B: From Appraisal to Physiological Responses). Such changes
could be expected to change the length and form of the vocal tract, likely affecting voice quality.
To jointly study intrinsic valence and goal congruence, Johnstone et al. (2005) modified a
popular computer game to present stimuli that were either congruent or incongruent with the
goal of winning and were accompanied by intrinsically positive or negative sounds. Acoustic
analysis of the standardized vocal utterances requested from the participants revealed that mean
energy, f0 level, utterance duration, and the proportion of the voiced part of the utterances varied
with goal congruence, whereas spectral energy distribution depended on intrinsic valence. Pitch
dynamics depended on the interaction between goal congruence and intrinsic valence.
This is true for many species of animals, including humans. In the case of high control, especially
in the presence of an aggressive tendency (e.g., in dominance fights), acoustic changes occur,
Annu. Rev. Psychol. 2019.70:719-745. Downloaded from www.annualreviews.org
including a rise in vocal energy (heard as loudness) and changes in formant structure (Briefer
2012, Ko et al. 2015). Low or no prospective control, especially when leading to disengagement
or resignation, should produce low sympathetic activity and lax muscle tone, which will likely
produce decreases in f0 and f0 range, vocal intensity, the energy of the harmonics, and the rate
of articulation. This pattern has repeatedly been found in measurements of speech expressing
sadness or disappointment (Goudbeek & Scherer 2010, Johnstone et al. 2001). When control is
moderate or uncertain, as in states of stress or high mental workload, a sympathetic response is
to be expected, resulting in high f0, high vocal intensity, and faster articulation ( Johnstone et al.
2001, Paulmann et al. 2016).
Johnstone et al. (2005) examined the effect of goal congruence and prospective control in a
computer task in which participants lost or gained points under two levels of difficulty. Results
showed interactions between gain or loss and difficulty: f0 was higher for losses than for gains
when difficulty was high but not when difficulty was low. Electroglottal measures showed shorter
glottal open times for losses than for gains (suggesting raised laryngeal tension). Skin conductance
indicated higher sympathetic arousal for losses than for gains, particularly when difficulty was high.
Multiple appraisal criteria. Apart from a few isolated attempts to manipulate appraisals, most
of the work on vocal expressions asked professional actors to portray sets of specific emotions
(portrayal studies) whereas some studies asked professional actors to enact scenarios designed to
elicit specific emotions (Stanislavski method). In a review of 104 studies on vocal expressions,
Juslin & Laukka (2003) concluded that the patterns of vocal parameters found across studies were
generally consistent with the theoretical predictions of Scherer (1986).
From these portrayal and enactment studies, one can also derive indirect evidence for the role
of appraisal criteria by using the profiles of these emotions on the major affective dimensions of
arousal, valence, and potency. The important role of arousal in determining vocal parameters in
the expression of emotion is well established. There is less evidence for the contribution of valence
and potency to vocal emotion expression. To investigate these dimensions further, Goudbeek &
Scherer (2010) used this indirect approach to analyze a large corpus of emotion enactments [the
Geneva Multimodal Emotional Portrayal (GEMEP) corpus (Bänziger & Scherer 2010, Bänziger
et al. 2012)] comprising 12 emotions that systematically vary with respect to valence, potency,
and arousal. The acoustic profiles found for specific emotions confirmed earlier findings obtained
from a similar corpus (Banse & Scherer 1996). In addition, Goudbeek & Scherer (2010) used
composite scores for acoustic parameters to determine the predictive power of these parameters
for the underlying dimensions of arousal, valence, and potency. They confirmed that arousal
dominated many vocal parameters, but they could also identify parameters that are specifically
related to valence and potency, in particular spectral balance and spectral noise.
Belyk & Brown (2014) suggested that weaker effects for valence and potency may depend on
differences between emotion families. In an enactment study, they divided emotional vocalizations
into motivational, moral, and aesthetic families. The results showed that valence interacted with
emotion family for both pitch and amplitude, suggesting that there may not be a common acoustic
code for valence across families of emotions.
Another reason for weak effects is that most research to date has assumed linear relationships
and has used multivariate linear statistics to analyze the data linking acoustic parameters to the
expression of different emotions. However, there is also a possibility of curvilinear relationships,
as in the case of positive valence being expressed by either low or high intensity depending on the
respective emotion family or on other factors such as task difficulty (see Johnstone et al. 2005) or
Access provided by PERI - ET - Addis Ababa University on 08/09/21. For personal use only.
et al. (2005), who used five vocally enacted emotions with weak and strong emotion intensity and
asked listeners to rate each enactment on arousal, valence, and potency. The portrayals were
also acoustically analyzed (e.g., by analyzing speech rate, voice intensity, f0, and spectral energy
distribution). Results showed (a) distinct rating patterns for arousal, valence, and potency for the
different emotions; (b) significant correlations of all three affective dimensions with several vocal
cues; and (c) the successful prediction of all affective dimensions except valence from the vocal cues.
Laukka & Elfenbein (2012) directly investigated the inference of appraisals from vocal expres-
sions. They had actors produce vocal enactments, after which they asked participants to rate the
situations that led to these enactments in terms of appraisal criteria (i.e., novelty, intrinsic valence,
goal congruence, urgency, agency, prospective control, and fairness). The perceived appraisal
profiles for the different emotions showed high inter-rater agreement and were generally in ac-
cordance with predictions based on appraisal theory. The appraisal ratings also correlated with a
variety of acoustic measures related to pitch, intensity, voice quality, and temporal characteristics
of the enactments. A recent follow-up study reported by Nordström et al. (2017) compared an
Australian and an Indian sample and found that appraisal criteria inferred from vocal expressions
were consistent across cultures.
Gestural Expression
The literature on gestural and other bodily expression is extremely scarce compared to that on
facial and vocal expression (but see de Gelder 2006, Lhommet & Marsella 2015). Dael et al.
(2012a,b) developed a body movement coding system, which allows one to describe visible activity
(gestures) in parts of the body, without interpreting this description in terms of emotions (similar
to FACS for the face). They applied the system to the GEMEP corpus mentioned above (which
contains multimodal videos of enactments) to investigate the extent to which gestural expression
patterns support different predictions in the literature. The results showed that few emotions were
characterized by a prototypical gesture pattern, and that most emotions were variably expressed by
multiple patterns, which may reflect configurations of appraisals and action tendencies. To follow
up on this, Dael et al. (2013) investigated how the expression of emotions that vary on arousal,
valence, and potency influences the perception of dynamic arm gestures. Participants rated the
arm movements of all emotional expressions from the GEMEP corpus with muted sound and
blurred faces on six spatiotemporal characteristics that were found to be related to emotion in
previous research (amount of movement, movement speed, force, fluency, size, and height or
vertical position). Arousal and potency were found to be strong determinants of the perception
of gestural dynamics, whereas the differences between positive and negative emotions were less
pronounced. These results confirm the importance of arm movement in communicating major
affective dimensions and show that gesture forms an integrated part of multimodal nonverbal
emotion communication.
plausibility or empirical support. Instead, the review focuses on the components of the emotion
process about which there is substantial agreement (although there may be disagreement about
Annu. Rev. Psychol. 2019.70:719-745. Downloaded from www.annualreviews.org
the roles assigned to each) and on the causal links among them. The purpose of this change of
emphasis is to bring new perspectives to the attention of empirical emotion researchers interested
in examining the underlying mechanisms, rather than a limited number of basic emotions as the
outcomes of these mechanisms.
The classic approach in emotion research has been to start with the end product of the emotion
process—individual emotions, as labeled by emotion words—and to try to empirically determine
the specificity of these emotions in terms of elicitors and response profiles across the different
components. This is problematic for at least two reasons. The first is that terms like anger, fear,
sadness, and joy are not homogeneous classes. The anger category may manifest itself as slight
irritation or violent rage, sadness as depressed resignation or intense despair, and fear as slight
worry or terrorized panic, with an infinite number of intermediate states for each. Moreover,
blends or mixtures of emotions are extremely common (Scherer & Meuleman 2013, Scherer et al.
2004). This extreme variability in the outcomes of emotion processes makes it exceedingly difficult
to organize systematic research by emotion labels. The second reason is that it is highly likely
that the labeling of an emotion episode occurs at the very end of the process, as an outcome
(see Figure 1). Ideally, hypothesis-guided research specifies the causal factors and their expected
interactions in determining the outcome. This is not the case in research that fixes the outcome
as a class, given that this type of research does not allow specifying theoretical predictions and is
thus reduced to associative measures of class differentiation. The research reviewed in this article
illustrates an alternative approach, which emphasizes the underlying mechanism of the emotion
process and the dynamic interactions among components. This approach should contribute to
building a coherent and plausible theoretical framework that can guide future research.
This does not imply that emotion labels should be banned from emotion research, but rather
that they should be placed where they belong: at the end of the emotion process. They should be
treated as one class of dependent variables, allowing for a large variety of type and intensity (see
the sidebar titled Model Testing in Semantic Sedimentation). This is of particular importance to
applied research in real-life settings, where the emotions encountered and studied are generally
more subtle, often constituting blends rather than strong prototypical exemplars. Adopting a
process approach, in addition to fostering hypothesis-guided designs, may also facilitate emotion
research in ecologically valid contexts.
This review discusses relationships among appraisals and three other components. Testing the
complete model shown in Figure 1, however, would require obtaining evidence for relationships
among all of the components. In this review, some patterns of convergence across multiple
components are clearly discernible. For instance, novelty appraisal generates an orientation
Since multicomponent data sets that allow sophisticated model fitting have not yet appeared, it is profitable to
explore alternative routes to examine the relationships among the components in the emotion process model. One
possibility is to exploit the sedimentation of real-world mechanisms in the semantics of emotion words. Scherer &
Fontaine (2018) used a large data set on the semantic facets of 24 major emotion terms in 28 world languages. This
data set provides very stable profiles on 142 facets covering all emotion components. Using principal component
analysis, the authors determined four appraisal factors (valence, novelty, prospective control, and other-agency),
five response component factors (approach versus rejection, assertive versus submissive, high versus low autonomic
arousal, control action tendency, and orientation) and four feeling factors (bad versus good, strong versus weak,
calm versus restless, and tired). The factor scores were used to test the process model in Figure 1 with theoretically
Access provided by PERI - ET - Addis Ababa University on 08/09/21. For personal use only.
guided hierarchical regression analyses. Two of the significant results were that bad (good) feeling is predicted
by valence and novelty appraisals, strongly mediated by approach (rejection) and assertive (submissive) response
Annu. Rev. Psychol. 2019.70:719-745. Downloaded from www.annualreviews.org
factors, and that weak (strong) feeling is predicted by valence and other-agency, strongly mediated by assertive
(submissive) tendencies and strengthened by control action tendencies.
tendency that seems to be manifested by eyebrow raising in the face. Positive valence appraisal
leads to the tendency to approach corresponding to smiling (AU12), and negative valence leads
to the tendency to avoid corresponding to frowning (AU4). Control and power appraisal results
in the tendency to be active (in the antisocial or prosocial sense), which apparently leads to loud,
powerful vocalizations (vocal energy). A combination of unexpectedness, negative valence, and
high control appraisals results in effort mobilization and activation that are likely to result in
massive changes in the ANS and SNS (especially heart rate and muscle tension). A major task for
future model fitting is the detailed examination of the interaction between the components and
the degree of coherence of the different components. This will require large-scale collaborations
among research teams with different methodological competencies.
Component coherence is also a central issue for the questions of (a) when an emotion episode
begins or what differentiates it from other, nonemotional, episodes and (b) when it becomes
conscious. A possible way to answer the first question is to define the start of an emotion episode
by an increased synchronization and coherence among the various components beyond a certain
threshold and its end by the coherence dropping below a threshold. The degree of coherence
might, in turn, be determined by the goal relevance of the event. The nature of the emotion
episode could be seen to be determined by the specific patterning over time of recursively generated
appraisal results. Such a kaleidoscopic arrangement of appraisal results can generate a near infinite
number of different emotion episodes without clear categorical boundaries (Scherer 2009a).
A possible way to answer the second question is to argue that a person will become aware of an
altered state after a second threshold of coherence has been reached. A plausible underlying mech-
anism is the central sensorimotor integration and representation of changes in all synchronized
components in the central nervous system (CNS), which receives massive projections from both
cortical and subcortical CNS structures (including proprioceptive feedback from the periphery)
and serves to monitor and regulate the degree of system coherence. Thus, the degree of coherence
among components might generate awareness, leading to the experience of feelings. People can
focus on micromomentary changes of feeling, but they usually become aware of them as temporal
chunks with phenomenal unity. These chunks might occur when a certain level of component
synchronization persists.
Research into component coherence and temporal chunking has barely begun. However, there
are increased interest and promising first investigations in both psychology and neuroscience.
In 2014, the journal Biological Psychology devoted a special issue, titled “Whither concordance?
Autonomic psychophysiology and the behaviors and cognitions of emotional responsivity,” to
this topic. This type of advanced research requires complex dynamic measurement of changes
in both the appraisal and response components, and sophisticated statistical and mathematical
tools are required for the analysis. Recently, Rügamer et al. (2017) developed and successfully
applied a new historical function-on-function regression model to measure the cross-correlation
of brain processes (electroencephalographic data) and muscle innervation (EMG data) that were
simultaneously recorded from participants while they were playing a computerized gambling task
(Gentsch et al. 2014). Our understanding of the nature of the emotion process is intimately tied
to further progress in this direction.
Access provided by PERI - ET - Addis Ababa University on 08/09/21. For personal use only.
Annu. Rev. Psychol. 2019.70:719-745. Downloaded from www.annualreviews.org
SUMMARY POINTS
1. We propose that research move from a discrete emotion approach to an emotion pro-
cess approach with an emphasis on the mechanisms and determinants underlying the
unfolding of emotion episodes, treating emotion labels as dependent variables.
2. The starting point for this movement is a process model (shared across many theories)
according to which emotion episodes comprise the following components: stimulus ap-
praisal, action tendencies, physiological responses, motor expression, experienced feeling,
and (in some cases) categorization and labeling.
3. Three relationships in the process model are selected for a review of the pertinent lit-
erature: the effect of different appraisal criteria on (a) action tendencies or behavior,
(b) physiological responses, and (c) motor expression (in face, voice, and gestures).
4. The review of research pertaining to relationships between appraisal and other compo-
nents is organized by the criteria of novelty or unexpectedness, (intrinsic and extrinsic)
valence, agency, control or power, and fairness.
5. Research on the influence of appraisal on action tendencies or behavior draws a complex
picture of how the various appraisal criteria influence action tendencies such as antisocial
versus prosocial tendencies, approach versus avoidance tendencies, and tendencies to be
active versus passive.
6. Work on physiological responses has searched for differences between discrete emo-
tions but has yielded few reliable patterns; in contrast, there is consistent evidence of
stable correlates of appraisal criteria on physiological correlates in the ANS and SNS, in
particular for cardiovascular parameters and muscle tension.
7. Facial expression has been a mainstay of emotion research since Darwin, yet it has been
difficult to show consistent differences in facial muscle activation patterns between dis-
crete emotions, even in actor portrayal studies; in contrast, experimental work on the
effects of appraisal manipulations shows well-replicated effects, especially for the effect
of valence on the zygomaticus (smile) and corrugator (frown) muscles.
8. Vocal expression has been relatively neglected in research, but recent literature using
advanced acoustic analyses shows many consistent response patterns, particularly linked
to control or power appraisals.
FUTURE ISSUES
1. Almost all of the work on emotion has been confined to the laboratory and, in many
expression studies, to actor-portrayed emotions, highlighting the need to branch out to-
ward more ecologically valid settings, despite the effort and cost involved. One promising
avenue might be the use of virtual reality environments and synthetic avatars that can be
easily manipulated.
2. While much of the work reviewed above has focused on the relationships between
pairs of components, several studies have measured two or more components, some of
which reveal coherence and others interaction effects, suggesting the importance of com-
plex multifactorial designs and advanced dynamic measurement rather than before–after
Access provided by PERI - ET - Addis Ababa University on 08/09/21. For personal use only.
designs.
Annu. Rev. Psychol. 2019.70:719-745. Downloaded from www.annualreviews.org
DISCLOSURE STATEMENT
The authors are not aware of any affiliations, memberships, funding, or financial holdings that
might be perceived as affecting the objectivity of this review.
ACKNOWLEDGMENTS
Preparation of this review was supported by (a) a European Research Council Advanced Grant
in the European Community’s seventh Framework Program under grant agreement 230331-
PROPEREMO (Production and perception of emotion: an affective sciences approach) to K.R.S.
and (b) research program G073317N of the Research Foundation—Flanders (FWO) and grant
C14/17/047 of the Research Council of KU Leuven to A.M.
LITERATURE CITED
Aue T, Flykt A, Scherer KR. 2007. First evidence for differential and sequential efferent effects of stimulus
relevance and goal conduciveness appraisal. Biol. Psychol. 74(3):347–57
Aue T, Scherer KR. 2008. Appraisal-driven somatovisceral response patterning: effects of intrinsic pleasantness
and goal conduciveness. Biol. Psychol. 79(2):158–64
Aue T, Scherer KR. 2011. Effects of intrinsic pleasantness and goal conduciveness appraisals on somatovisceral
responding: somewhat similar, but not identical. Biol. Psychol. 86(1):65–73
Bachorowski J-A, Owren MJ. 1995. Vocal expression of emotion: acoustic properties of speech are associated
with emotional intensity and context. Psychol. Sci. 6(4):219–24
Bachorowski J-A, Smoski MJ, Owen MJ. 2001. The acoustic features of human laughter. J. Acoust. Soc. Am.
110(3 Pt. 1):1581–97
Banse R, Scherer KR. 1996. Acoustic profiles in vocal emotion expression. J. Personal. Soc. Psychol. 70(3):614–36
Bänziger T, Mortillaro M, Scherer KR. 2012. Introducing the Geneva Multimodal expression corpus for
experimental research on emotion perception. Emotion 12(5):1161–79
Bänziger T, Scherer KR. 2010. Introducing the Geneva Multimodal Emotion Portrayal (GEMEP) corpus.
In Blueprint for Affective Computing: A Sourcebook, ed. KR Scherer, T Bänziger, EB Roesch, pp. 271–94.
Oxford, UK: Oxford Univ. Press
Barlett C, Witkower Z, Mancini C, Saleem M. 2016. Breaking the link between provocation and aggression:
the role of mitigating information. Aggress. Behav. 42(6):555–62
Barlett CP. 2013. Excuses, excuses: a meta-analytic review of how mitigating information can change aggression
Access provided by PERI - ET - Addis Ababa University on 08/09/21. For personal use only.
Bediou B, Scherer KR. 2014. Egocentric fairness perception: emotional reactions and individual differences
in overt responses. PLOS ONE 9(2):e88432
Belyk M, Brown S. 2014. The acoustic correlates of valence depend on emotion family. J. Voice 28(4):523.e9–18
Berkowitz L, Harmon-Jones E. 2004. Toward an understanding of the determinants of anger. Emotion
4(2):107–30
Bierbrauer F, Ockenfels A, Pollak A, Rückert D. 2017. Robust mechanism design and social preferences.
J. Public Econ. 149:59–80
Bossuyt E, Moors A, De Houwer J. 2014. Exploring the relations between regret, self-agency, and the tendency
to repair using experimental methods and structural equation modeling. Psychol. Rec. 64(4):841–57
Botvinick MM. 2007. Conflict monitoring and decision making: reconciling two perspectives on anterior
cingulate function. Cogn. Affect. Behav. Neurosci. 7(4):356–66
Briefer EF. 2012. Vocal expression of emotions in mammals: mechanisms of production and evidence. J. Zool.
288(1):1–20
Brosch T, Scherer KR, Grandjean DM, Sander D. 2013. The impact of emotion on perception, attention,
memory, and decision-making. Swiss Med. Wkly. 143:w13786
Carlsmith JM, Gross AE. 1969. Some effects of guilt on compliance. J. Personal. Soc. Psychol. 11(3):232–39
Chalabaev A, Major B, Cury F, Sarrazin P. 2009. Physiological markers of challenge and threat mediate the
effects of performance-based goals on performance. J. Exp. Soc. Psychol. 45(4):991–94
Chen S, Lee-Chai AY, Bargh JA. 2001. Relationship orientation as a moderator of the effects of social power.
J. Personal. Soc. Psychol. 80(2):173–87
Cherek D, Spiga R, Steinberg J, Kelly T. 1990. Human aggressive responses maintained by avoidance or
escape from point loss. J. Exp. Anal. Behav. 53(2):293–303
Civai C. 2013. Rejecting unfairness: emotion-driven reaction or cognitive heuristic? Front. Hum. Neurosci.
7:126
Dael N, Goudbeek M, Scherer KR. 2013. Perceived gesture dynamics in nonverbal expression of emotion.
Perception 42(6):642–57
Dael N, Mortillaro M, Scherer KR. 2012a. The body action and posture coding system (BAP): development
and reliability. J. Nonverbal Behav. 36(2):97–121
Dael N, Mortillaro M, Scherer KR. 2012b. Emotion expression in body action and posture. Emotion
12(5):1085–101
de Gelder B. 2006. Towards the neurobiology of emotional body language. Nat. Rev. Neurosci. 7(3):242–49
De Houwer J, Moors A. 2010. Implicit measures: similarities and differences. In Handbook of Implicit Social
Cognition: Measurement, Theory, and Applications, ed. B Gawronski, K Payne, pp. 176–93. New York:
Guilford Press
Delplanque S, Grandjean D, Chrea C, Coppin G, Aymard L, et al. 2009. Sequential unfolding of novelty
and pleasantness appraisals of odors: evidence from facial electromyography and autonomic reactions.
Emotion 9(3):316–28
Denson TF, Spanovic M, Miller N. 2009. Cognitive appraisals and emotions predict cortisol and immune
responses: a meta-analysis of acute laboratory social stressors and emotion inductions. Psychol. Bull.
135(6):823–53
Dignath D, Kiesel A, Eder AB. 2015. Flexible conflict management: conflict avoidance and conflict adjustment
in reactive cognitive control. J. Exp. Psychol. Learn. Mem. Cogn. 41(4):975–88
Donnerstein E, Wilson DW. 1976. Effects of noise and perceived control on ongoing and subsequent aggres-
sive behavior. J. Personal. Soc. Psychol. 34(5):774–81
Eder AB, Rothermund K. 2008. When do motor behaviors (mis)match affective stimuli? An evaluative coding
view of approach and avoidance reactions. J. Exp. Psychol. 137:262–81
Ekman P. 1997. What we have learned by measuring facial behavior. In What the Face Reveals, ed. P Ekman,
EI Rosenberg, pp. 469–85. Oxford, UK: Oxford Univ. Press
Ekman P, Friesen WV. 1978. Facial Action Coding System: A Technique for the Measurement of Facial Movement.
Palo Alto, CA: Consult. Psychol. Press
Access provided by PERI - ET - Addis Ababa University on 08/09/21. For personal use only.
Ellsworth PC, Scherer KR. 2003. Appraisal processes in emotion. In Handbook of Affective Sciences, ed. RJ
Annu. Rev. Psychol. 2019.70:719-745. Downloaded from www.annualreviews.org
Davidson, KR Scherer, HH Goldsmith, pp. 572–95. Oxford, UK: Oxford Univ. Press
Fast NJ, Chen S. 2009. When the boss feels inadequate: power, incompetence, and aggression. Psychol. Sci.
20:1406–13
Fleischmann A, Lammers J, Conway P, Galinsky AD. 2018. Paradoxical effects of power on moral thinking:
why power both increases and decreases deontological and utilitarian moral decisions. Soc. Psychol. Personal.
Sci. In press
Frijda N, Scherer KR. 2009. Emotion definitions (psychological perspectives). In Oxford Companion to Emotion
and the Affective Sciences, ed. D Sander, KR Scherer, pp. 142–44. Oxford, UK: Oxford Univ. Press
Frijda NH. 1986. The Emotions. Cambridge, UK: Cambridge Univ. Press
Frijda NH, Kuipers P, ter Schure E. 1989. Relations among emotion, appraisal, and emotional action readiness.
J. Personal. Soc. Psychol. 57:212–28
Gaab J, Rohleder N, Nater UM, Ehlert U. 2005. Psychological determinants of the cortisol stress response:
the role of anticipatory cognitive appraisal. Psychoneuroendocrinology 30(6):599–610
Galinsky AD, Gruenfeld DH, Magee JC. 2003. From power to action. J. Personal. Soc. Psychol. 85:453–66
Galvin BM, Randel AE, Collins BJ, Johnson RE. 2018. Changing the focus of locus (of control): a targeted
review of the locus of control literature and agenda for future research. J. Organ. Behav. 38:820–33
Gentsch K, Grandjean D, Scherer KR. 2014. Coherence explored between emotion components: evidence
from event-related potentials and facial electromyography. Biol. Psychol. 98:70–81
Gentsch K, Grandjean D, Scherer KR. 2015a. Appraisals generate specific configurations of facial muscle
movements in a gambling task: evidence for the component process model of emotion. PLOS ONE
10(8):e0135837
Gentsch K, Grandjean D, Scherer KR. 2015b. Temporal dynamics and potential neural sources of goal
conduciveness, control, and power appraisal. Biol. Psychol. 112:77–93
Goudbeek M, Scherer KR. 2010. Beyond arousal: valence and potency/control in the vocal expression of
emotion. J. Acoust. Soc. Am. 128(3):1322–36
Gramer M, Supp N. 2014. Social support and prolonged cardiovascular reactivity: the moderating influence
of relationship quality and type of support. Biol. Psychol. 101:1–8
Gray JA, McNaughton N. 1996. The neuropsychology of anxiety: reprise. Nebr. Symp. Motiv. 43:61–134
Guinote A. 2017. How power affects people: activating, wanting, and goal seeking. Annu. Rev. Psychol. 68:353–
81
Güth W, Kocher MG. 2014. More than thirty years of ultimatum bargaining experiments: motives, variations,
and a survey of the recent literature. J. Econ. Behav. Organ. 108:396–409
Hamm AO, Schupp HT, Weike AI. 2003. Motivational organization of emotions: autonomic changes, cortical
responses and reflex modulation. In Handbook of Affective Sciences, ed. RJ Davidson, KR Scherer, HH
Goldsmith, pp. 187–211. Oxford, UK: Oxford Univ. Press
Harvey A, Nathens AB, Bandiera G, LeBlanc VR. 2010. Threat and challenge: cognitive appraisal and stress
responses in simulated trauma resuscitations. Med. Educ. 44(6):587–94
Heller AS, Lapate RC, Mayer KE, Davidson RJ. 2014. The face of negative affect: Trial-by-trial corrugator
responses to negative pictures are positively associated with amygdala and negatively associated with
ventromedial prefrontal cortex activity. J. Cogn. Neurosci. 26(9):2102–10
Hiroto DS. 1974. Locus of control and learned helplessness. J. Exp. Psychol. 102(2):187–93
Johnstone T, van Reekum CM, Hird K, Kirsner K, Scherer KR. 2005. Affective speech elicited with a computer
game. Emotion 5(4):513–18
Johnstone T, van Reekum CM, Scherer KR. 2001. Vocal correlates of appraisal processes. In Scherer et al.
2001, pp. 271–84
Juslin PN, Laukka P. 2003. Communication of emotions in vocal expression and music performance: different
channels, same code? Psychol. Bull. 129(5):770–814
Juslin PN, Scherer KR. 2005. Vocal expression of affect. In The New Handbook of Methods in Nonverbal Behavior
Research, ed. JA Harrigan, R Rosenthal, K Scherer, pp. 65–135. Oxford, UK: Oxford Univ. Press
Kemeny ME. 2009. Psychobiological responses to social threat: evolution of a psychological model in psy-
Access provided by PERI - ET - Addis Ababa University on 08/09/21. For personal use only.
Ko SJ, Sadler MS, Galinsky AD. 2015. The sound of power: conveying and detecting hierarchical rank through
voice. Psychol. Sci. 26(1):3–14
Konecni VJ. 1972. Some effects of guilt on compliance: a field replication. J. Personal. Soc. Psychol. 23(1):30–32
Kozlik J, Neumann R, Lozo L. 2015. Contrasting motivational orientation and evaluative coding accounts:
on the need to differentiate the effectors of approach/avoidance responses. Front. Psychol. 6:563
Kreibig SD. 2010. Autonomic nervous system activity in emotion: a review. Biol. Psychol. 84(3):394–421
Krieglmeyer R, De Houwer J, Deutsch R. 2013. On the nature of automatically triggered approach-avoidance
behavior. Emot. Rev. 5(3):280–84
Krieglmeyer R, Deutsch R. 2013. Approach does not equal approach: Angry facial expressions evoke approach
only when it serves aggression. Soc. Psychol. Personal. Sci. 4(5):607–14
Kulik JA, Brown R. 1979. Frustration, attribution of blame, and aggression. J. Exp. Soc. Psychol. 15(2):183–94
Kuppens P, Van Mechelen I, Smits DJM, De Boeck P. 2003. The appraisal basis of anger: specificity, necessity,
and sufficiency of components. Emotion 3(3):254–69
Lammers J, Galinsky AD, Gordijn EH, Otten S. 2008. Illegitimacy moderates the effects of power on approach.
Psychol. Sci. 19(6):558–64
Lanctôt N, Hess U. 2007. The timing of appraisals. Emotion 7(1):207–12
Lang PJ. 2014. Emotion’s response patterns: the brain and the autonomic nervous system. Emot. Rev. 6(2):93–
99
Larsen JT, Norris CJ, Cacioppo JT. 2003. Effects of positive and negative affect on electromyographic activity
over zygomaticus major and corrugator supercilii. Psychophysiology 40(5):776–85
Laukka P, Elfenbein HA. 2012. Emotion appraisal dimensions can be inferred from vocal expressions. Soc.
Psychol. Personal. Sci. 3(5):529–36
Laukka P, Juslin P, Bresin R. 2005. A dimensional approach to vocal expression of emotion. Cogn. Emot.
19(5):633–53
Lazarus RS, Folkman S. 1984. Stress, Appraisal, and Coping. New York: Free Press
Lee DH, Susskind JM, Anderson AK. 2013. Social transmission of the sensory benefits of eye widening in fear
expressions. Psychol. Sci. 24(6):957–65
Lhommet M, Marsella SC. 2015. Expressing emotion through posture and gesture. In The Oxford Handbook
of Affective Computing, ed. RA Calvo, SK D’Mello, J Gratch, A Kappas, pp. 273–85. Oxford, UK: Oxford
Univ. Press
Lole L, Gonsalvez CJ, Blaszczynski A, Clarke AR. 2012. Electrodermal activity reliably captures physiological
differences between wins and losses during gambling on electronic machines. Psychophysiology 49(2):154–
63
Löw A, Lang PJ, Smith JC, Bradley MM. 2008. Both predator and prey: emotional arousal in threat and
reward. Psychol. Sci. 19(9):865–73
Martinez LM, Zeelenberg M, Rijsman JB. 2011. Behavioural consequences of regret and disappointment in
social bargaining games. Cogn. Emot. 25(2):351–59
Matarazzo O, Pizzini B, Greco C, Carpentieri M. 2016. Effects of a chance task outcome on the offers in
the ultimatum game: the mediation role of emotions. In 7th IEEE International Conference on Cognitive
Infocommunications, 16–18 Oct., pp. 295–300. New York: IEEE
McCloskey MS, Berman ME, Coccaro EF. 2005. Providing an escape option reduces retaliatory aggression.
Aggress. Behav. 31(3):228–37
Mehu M, Scherer KR. 2015. Emotion categories and dimensions in the facial communication of affect: an
integrated approach. Emotion 15(6):798–811
Moors A. 2009. Theories of emotion causation: a review. Cogn. Emot. 23(4):625–62
Moors A. 2017. The integrated theory of emotional behavior follows a radically goal-directed approach.
Psychol. Inq. 28(1):68–75
Moors A, Boddez Y, De Houwer J. 2017. The power of goal-directed processes in the causation of emotional
and other actions. Emot. Rev. 9(4):310–18
Moors A, De Houwer J. 2006. Automaticity: a theoretical and conceptual analysis. Psychol. Bull. 132:297–
Access provided by PERI - ET - Addis Ababa University on 08/09/21. For personal use only.
326
Moors A, Fischer M. 2018. Demystifying the role of emotion in behavior: toward a goal-directed account.
Annu. Rev. Psychol. 2019.70:719-745. Downloaded from www.annualreviews.org
Scherer KR. 2001. Appraisal considered as a process of multilevel sequential checking. In Scherer et al. 2001,
pp. 92–120
Scherer KR. 2009a. The dynamic architecture of emotion: evidence for the component process model. Cogn.
Emot. 23(7):1307–51
Scherer KR. 2009b. Emotion theories and concepts (psychological perspectives). In Oxford Companion to
Emotion and the Affective Sciences, ed. D Sander, KR Scherer, pp. 145–49. Oxford, UK: Oxford Univ.
Press
Scherer KR. 2013. Emotion in action, interaction, music, and speech. In Language, Music, and the Brain: A
Mysterious Relationship, ed. MA Arbib, pp. 107–39. Cambridge, MA: MIT Press
Scherer KR, Clark-Polner E, Mortillaro M. 2011. In the eye of the beholder? Universality and cultural
specificity in the expression and perception of emotion. Int. J. Psychol. 46(6):401–35
Scherer KR, Ellgring H. 2007. Are facial expressions of emotion produced by categorical affect programs or
dynamically driven by appraisal? Emotion 7(1):113–30
Access provided by PERI - ET - Addis Ababa University on 08/09/21. For personal use only.
Scherer KR, Fontaine JRF. 2018. The semantic structure of emotion words across languages is consistent
with componential appraisal models of emotion. Cogn. Emot. In press
Annu. Rev. Psychol. 2019.70:719-745. Downloaded from www.annualreviews.org
Scherer KR, Meuleman B. 2013. Human emotion experiences can be predicted on theoretical grounds: evi-
dence from verbal labeling. PLOS ONE 8(3):e58166
Scherer KR, Mortillaro M, Mehu M. 2017. Facial expression is driven by appraisal and generates appraisal
inference. In The Science of Facial Expression, ed. JM Fernández-Dols, JA Russell, pp. 353–73. Oxford, UK:
Oxford Univ. Press
Scherer KR, Mortillaro M, Rotondi I, Sergi I, Trznadel S. 2018. Appraisal-driven facial actions as building
blocks for emotion inference. J. Personal. Soc. Psychol. 114(3):358–79
Scherer KR, Peper M. 2001. Psychological theories of emotion and neuropsychological research. In Hand-
book of Neuropsychology, Vol. 5: Emotional Behavior and Its Disorders, ed. F Boller, J Grafman, pp. 17–48.
Amsterdam: Elsevier
Scherer KR, Schorr A, Johnstone T, eds. 2001. Appraisal Processes in Emotion: Theory, Methods, Research. Oxford,
UK: Oxford Univ. Press
Scherer KR, Wranik T, Sangsue J, Tran V, Scherer U. 2004. Emotions in everyday life: probability of
occurrence, risk factors, appraisal and reaction pattern. Soc. Sci. Inform. 43(4):499–570
Seligman MEP. 1975. Helplessness: On Depression, Development, and Death. San Francisco: Freeman
Silvia PJ, Nusbaum EC, Eddington KM, Beaty RE, Kwapil TR. 2014. Effort deficits and depression: the
influence of anhedonic depressive symptoms on cardiac autonomic activity during a mental challenge.
Motiv. Emot. 38(6):779–89
Smith CA, Kirby LD. 2009. Relational antecedents of appraised problem-focused coping potential and its
associated emotions. Cogn. Emot. 23(3):481–503
Smith PK, Bargh JA. 2008. Nonconscious effects of power on basic approach and avoidance tendencies. Soc.
Cogn. 26(1):1–24
Stajkovic AD, Luthans F. 1998. Self-efficacy and work-related performance: a meta-analysis. Psychol. Bull.
124:240–61
Stekelenburg JJ, van Boxtel A. 2002. Pericranial muscular, respiratory, and heart rate components of the
orienting response. Psychophysiology 39(6):707–22
Sumpf M, Jentschke S, Koelsch S. 2015. Effects of aesthetic chills on a cardiac signature of emotionality. PLOS
ONE 10(6):e0130117
Talsma K, Schüz B, Schwarzer R, Norris K. 2018. I believe, therefore I achieve (and vice versa): a meta-analytic
cross-lagged panel analysis of self-efficacy and academic performance. Learn. Individ. Differ. 61:136–
50
Tomaka J, Blascovich J, Kelsey RM, Leitten CL. 1993. Subjective, physiological, and behavioral effects of
threat and challenge appraisal. J. Personal. Soc. Psychol. 65(2):248–60
Tomaka J, Blascovich J, Kibler J, Ernst JM. 1997. Cognitive and physiological antecedents of threat and
challenge appraisal. J. Personal. Soc. Psychol. 73(1):63–72
Triscoli C, Croy I, Olausson H, Sailer U. 2017. Touch between romantic partners: Being stroked is more
pleasant than stroking and decelerates heart rate. Physiol. Behav. 177:169–75
van Damme E, Binmore KG, Roth AE, Samuelson L, Winter E, et al. 2014. How Werner Güth’s ultimatum
game shaped our understanding of social behavior. J. Econ. Behav. Organ. 108:292–318
van Reekum CM, Johnstone T, Banse R, Etter A, Wehrle T, Scherer KR. 2004. Psychophysiological responses
to appraisal dimensions in a computer game. Cogn. Emot. 18(5):663–88
Weinstein SE, Quigley KS. 2006. Locus of control predicts appraisals and cardiovascular reactivity to a novel
active coping task. J. Personal. 74(3):911–31
Wortman CB, Brehm JW. 1975. Responses to uncontrollable outcomes: an integration of reactance theory
and the learned helplessness model. In Advances in Experimental Social Psychology, Vol. 8, ed. L Berkowitz,
pp. 277–336. New York: Academic
Wu L, Winkler MH, Andreatta M, Hajcak G, Pauli P. 2012. Appraisal frames of pleasant and unpleasant
pictures alter emotional responses as reflected in self-report and facial electromyographic activity. Int. J.
Psychophysiol. 85(2):224–29
Zeelenberg M, Beattie J. 1997. Consequences of regret aversion 2: additional evidence for effects of feedback
Access provided by PERI - ET - Addis Ababa University on 08/09/21. For personal use only.
Annual Review of
Psychology
Contents
Access provided by PERI - ET - Addis Ababa University on 08/09/21. For personal use only.
vi
PS70_FrontMatter ARI 10 November 2018 11:59
Contents vii
PS70_FrontMatter ARI 10 November 2018 11:59
A New Era of HIV Risk: It’s Not What You Know, It’s Who You Know
(and How Infectious)
Andrew C. Cortopassi, Redd Driver, Lisa A. Eaton, and Seth C. Kalichman p p p p p p p p p p p 673
Stress and Obesity
A. Janet Tomiyama p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 703
The Emotion Process: Event Appraisal and Component Differentiation
Klaus R. Scherer and Agnes Moors p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 719
How to Do a Systematic Review: A Best Practice Guide for Conducting
and Reporting Narrative Reviews, Meta-Analyses, and Meta-Syntheses
Andy P. Siddaway, Alex M. Wood, and Larry V. Hedges p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 747
Access provided by PERI - ET - Addis Ababa University on 08/09/21. For personal use only.
Annu. Rev. Psychol. 2019.70:719-745. Downloaded from www.annualreviews.org
Indexes
Errata
viii Contents